oral history interview with robert frater cardiothoracic

Oral History Interview with

Robert Frater

Cardiothoracic Surgeon

St. Jude’s Medical Center

FDA Oral History Program

Final Edited Transcript

May 2003

Robert Frater Oral History 1

Table of Contents

Oral History Abstract ...................................................................................................................... 2

Keywords ........................................................................................................................................ 2

Citation Instructions ........................................................................................................................ 2

Interviewer Biography ..................................................................................................................... 3

FDA Oral History Program Mission Statement .............................................................................. 3

Statement on Editing Practices ....................................................................................................... 3

Index ............................................................................................................................................... 4

Interview Transcript ........................................................................................................................ 5


Oral History Abstract

This interview was conducted in an effort to collect background information on the development

of cardiothoracic surgery and heart valve design and surgical implantation. Dr. Frater was a

pioneer in the development of heart valve replacement surgery and the technical design of heart

valves, who worked at the University of Capetown Medical School, the University of Minnesota

and the Mayo Clinic from the 1950s to the 1980s.

Keywords

heart valve replacement surgery; mitral valve prosthesis; cardiac surgery; medical device

standards

Citation Instructions

This interview should be cited as follows:

“Robert Frater Oral History Interview,” History Office, U.S. Food and Drug Administration,

Department of Health and Human Services, May 2003.


Interviewer Biography

Suzanne Junod, Ph.D. is an historian in the FDA History Office at the U.S. Food and Drug

Administration. Soon after beginning her career at FDA in 1984, Suzanne helped to organize the

FDA History Office. She is a subject matter expert in FDA history and her scholarly writings

have been published in the Food, Drug, and Cosmetic Law Journal, the Journal of Federal

History, and the Journal of the History of Medicine and Allied Sciences, as well as edited

compilations. She is an active officer in the Society for History in the Federal Government. She

earned her Ph.D. at Emory University in Atlanta, where she studied under James Harvey

Young.

FDA Oral History Program Mission Statement

The principal goal of FDA’s OHP is to supplement the textual record of the Agency’s history to

create a multi-dimensional record of the Agency’s actions, policies, challenges, successes, and

workplace culture. The OHP exists to preserve institutional memory, to facilitate scholarly and

journalistic research, and to promote public awareness of the history of the FDA. Interview

transcripts are made available for public research via the FDA website, and transcripts as well as

audio recordings of the interviews are deposited in the archives of the National Library of

Medicine. The collection includes interviews with former FDA employees, as well as members

of industry, the academy and the legal and health professions with expertise in the history of

food, drug and cosmetic law, policy, commerce and culture. These oral histories offer valuable

first-person perspectives on the Agency’s work and culture, and contribute otherwise

undocumented information to the historical record.

Statement on Editing Practices

It is the policy of the FDA Oral History Program to edit transcripts as little as possible, to ensure

that they reflect the interviewee’s comments as accurately as possible. Minimal editing is

employed to clarify mis-starts, mistakenly conveyed inaccurate information, archaic language,

and insufficiently explained subject matter. FDA historians edit interview transcripts for copy

and content errors. The interviewee is given the opportunity to review the transcript and suggest

revisions to clarify or expand on interview comment, as well as to protect their privacy, sensitive

investigative techniques, confidential agency information, or trade secrets.


Index

anti-communism, 29

aortic valve prosthesis

developmen tof, 26

Artificial Valves for Cardiac Surgery, 19

Barnard, Chris, 23, 26

Beall valve, 38

Bjork, Viking, 36

Bjork-Shiley valve, 35, 36, 37

Brownwald, Nina, 20

cardiac surgery

Gibbon machine, 9, 10

career

Mayo Clinic, 7

Cinematography, 15

Clark, Richard (Dick), 27

Copper Medics valve, 40

Curtin, John, 9, 17, 23

early cardiac open-heart cases, 22

education, 5

cardiac surgery, 8

Master's thesis, 15

Mayo Clinic fellowship, 11

Edwards Laboratory, 30

Edwards valve, 33, 34

Edwards, Lowell, 21, 29

embolism, 35

Fitell, Bruce, 33, 35, 36, 37

Hancock 2 valve, 33

heart valve standards, 40

Hepatitis, 24

homograft, 27

Ionescu 2 valve, 32, 33

Ionescu, Marian, 31, 32, 33

Ionescu-Shiley valve, 28, 29, 30

laboratory animals, 34

Medical Device Amendments, 37

Medtronic Hall valve, 38

mitral valve prosthesis

development of, 11, 12, 13, 16, 26

first patient, 17

wear testing, 27, 30

mitral valve prothesis

manufacturer, 21

Mitroflow valve, 33

Morrow, Glen, 20

pericardium procurement, 31

personal history

youth, 5

porcine valve, 27, 28, 33

Shiley Company, 32, 33, 35, 36

Shiley, Donald, 29, 31, 37

St. Jude valve, 38

Starr valve, 21, 38

Starr, Al, 22

stents

pericardial v. cloth covering, 32

thrombosis, 35

University of Capetown, 23, 24

valve replacement surgery

methods, 18

sutures, 19, 38


Interview Transcript

SJ: I’m Suzanne White Junod, and we’re here today with Robert Frater, a cardiothoracic

surgeon. He now works for St. Jude Medical. We’re at the American Association for

Cardiothoracic Surgery in May 2003. Dr. Frater, can you tell us a little about your background?

RF: Well, I grew up in South Africa, where my family had been for many generations,

although my mother, in fact, was born in England. She and my father had met at the Mayo

Clinic, where they were both doing postgraduate work, he from South Africa, she from the U.K.,

and came back to South Africa, and I grew up essentially knowing I was going to be a doctor. As

soon as I got past the engine-driver or locomotive-engineer stage, it was going to be medicine.

So I went to the University of Capetown Medical School, entering it, as is the case in

South Africa, straight after high school at the age of eighteen. And we had in the University of

Capetown at that time a quite remarkable Division of Cardiology. There was a doctor called

Velva Freier, who led the division. Everybody who was an internist or a physician [unclear] for

cardiologists in those days, but he had started to specialize, and he had working with him Louis

Vogelpohl or Louie Vogelpohl, who had been to London to learn about phonocardiography from

Paul Wood. He himself had been to America and learned about dye dilution curves at the Mayo

Clinic, and we had a Robert Gertz, who had an obituary or a biography in the Annals of Thoracic

Surgery a couple of years ago, a remarkable man who’d come out of Germany and then to

Scotland, and then finally to South Africa, where he had set up a department of surgical research.

He was a person with an interest in surgery. He never really trained in it fully, but he did,

for example – I’ll give you the atmosphere I went to medical school in – he did triptosomography

of a standard that I’ve never seen equaled before. I won’t waste time or spend time on the very


ingenious way he did that. He had a thoracoscopic sympathectomy operation which he published

in 1944 for Reynaud’s phenomenon. And he had developed angiography of peripheral vessels in

the mid-1940s with a rapid cassette changer he invented himself, which subsequently was more

or less taken over by G.E. without any exchange of money or anything.

And so we had in the University of Capetown Medical School, when I was a student,

phonocardiography, dye dilution analyses, the usual pressures and cardiac outputs, and

angiography as tools for investigation. And as students, we took it for granted that was the

normal. That was in fact very unusual in most parts of the world. We listened to a lecture on

100 successful cardiac catheterizations with angiography in 1949, during our physiology classes.

So we all knew an immense amount of cardiology. But what were we to do with it, or what was

I to do with it as somebody whose father was a surgeon and who wanted to be a surgeon himself?

I could have done closed mitral, patent ductus arteriosis, co-optation, and Blalock shunt. Those

were the only four operations to do.

And so I got a fellowship appointment at the Mayo Clinic.

I went first for a year to London, and young South Africans would do that in those days

because we figured we had to get out of the country. We were down at the tip of Africa, isolated

from the rest of the world, and we thought we had to go learn something from other places if we

were ambitious at all. And so I went to the Royal College of Surgeons and did their basic science

course. And while I was there, a lecture was given by Crawford, a Swedish cardiac surgeon, who

described doing an atrial septal defect by dissecting the intra-atrial groove, passing a great big

needle blind, with a finger guiding it, through the right atrial appendage, around the edges of an

atrial [unclear] and out again, and cobbling the whole thing up to close an ASD. So I said to

myself, “My golly, there’s more to it than the operations.” I thought, “You can get inside the


heart.” And at that particular Thursday afternoon routine weekly lecture at the Royal College of

Surgeons, who did I meet but Henry Suttar.

SJ: What year is this?

RF: This is 1955, 1954. I met Henry Suttar, who had done the first digital mitral

commissurotomy in 1926, very [unclear]. After all, ’26 to ’54, that wasn’t a tremendously long

time after he’d done that operation, but he was philosophical about it. By now, Bailey, Harken,

and Brock had started commissurotomy again, and successfully, and he had had his one great

chance and had been denied it by the cardiologist.

Anyway, I then did a casualty job, and my wife and I and our newborn child went off to

the Mayo Clinic, and it was pretty damned depressing because the Mayo Clinic put us into

medicine. Even though I was a would-be surgeon, they said, “Well, you need to learn more

medicine if you’re going to be a surgeon.” To my mind, this was fairly straight forward

exploitation. They needed what they called, you know, they needed workers to do this and to

cover the medical floors.

And I was put on a diagnostic service, and I had modest hospital duties. It was called

medical urology, which meant that we saw all the patients coming to the urology service, worked

them up for the surgeons, and when they came to the hospital, we worked them up to be ready for

the next week’s surgery. And I found myself making diagnoses in cardiology which were quite

foreign to these trained general physicians in charge of this diagnostic service. They had not

understood the kinds of concepts that we took for granted.

I saw a young woman with very loud pulmonary systolic click. I remember it only


because of the excitement it caused amongst the attendings, who thought, “Who’s this little jerk

telling us this is a case of pulmonary hypertension? It may be due to chronic pulmonary

embolism from her legs. Although she has no signs of it, she’s had several children. Or it may

be a primary pulmonary hypertension.” And they said it’s pulmonary hypertension, you can feel

it, and so on. And of course I was right when they got the patient ultimately. But I realized I still

wanted to do cardiology, cardiac surgery more than anything.

And it was about a month after I’d gotten there – this is now September 1955 – I was

getting pretty depressed being paid $1,500 a year and miles away from everywhere, 9,000 miles

away from home, and I said, “I’ve just got to get into an operating . . .”

Saturday, I’d finished working up the prostatectomies for the next week, and I said, “I just

have to get into an operating room.” So on Saturday afternoon, in the old Methodist Hospital in

Rochester, Minnesota, I went into the operating room, suite, and I saw a heart surgery. Oh, my

God. And I went into the gallery, and the galleries were such, if that was the operating table, the

gallery was that high, and you just leaned over and looked right down into the operation. You

had a paper mask on and a gown which you’d picked up outside. I don’t think we even had to

put a cap on, if I remember correctly.

And John Curtin [could be spelled Kirklin] was the surgeon. A fellow called Harry

Harshbarger, who had been a tackle for Northwestern, was the assistant. He was so damn big, I

really had to lean over far to see the operation. And a fellow called Bob DeVue was giving the

anesthesia. And there was this big stainless steel machine with a vertical portion down which

blood was trickling on a screen. And the blood was then moved towards the patient with a roller

pump put through a plastic pipe and into the patient, and coming from the patient there was

another of these lines with another roller pump on it. And what they were doing, this machine


was quite extraordinary. It was an IBM-designed machine. It was a Gibbon machine which

actually had a sensor – and this I only found out afterwards – on the central venous pressure and

it altered the venous return by speeding up or slowing down according to the central venous

pressure. And running the pump was Jeremy Swann of Swann gas fame; David Donald, a

Scottish veterinarian; and a physiologist called Edwards who – not Jess Edwards the pathologist,

but another Edwards who developed the G suit during the war for flyers; there was a human

centrifuge in Rochester. And I realized they’re operating on this kid’s heart and they’re using a

machine to support the kid’s life.

Now, I’d heard of the possibility, because while I was a student in Capetown, we actually

had a lecture from a visiting Dutchman called Jon Blut, which means young blood in English,

and who’d described efforts to develop a heart-lung machine that he’d been making. And of

course I subsequently found out this long and interesting history of the development of heart-lung

machines, but at that moment it was . . .

The next day, I said to my wife, “You know, we’re going to stay longer in Rochester than

we’d planned, if you don’t mind, because I’m going into cardiac surgery.” And she said okay.

And so I went along to the fellowship office and said, “I’d like to apply for the

cardiothoracic program, which was then an add-on to the general, it was all mixed in together in

terms of rotations.

Anyway, I then had to do a thesis. There were two cardiac surgeons at the Mayo Clinic at

the time: John Curtin and Bucky Ellis, and Curtin was a young man. The position had been

offered to Claggett, who was the senior thoracic surgeon, but he said, “I’m too old.” He was in

his late forties, I suppose. And he said, “Give it to John. He’s just spent a year in Harvard at the

Massachusetts General – or was it the Brigham – working with Gross, and he knows more about


it than I do. So he developed it.

Maybe I should say a few words about the machine.

He told me that when you wanted to go in, he did the usual thing. You go to the

literature, see what’s there, and he found there was a machine that Dennis had used in New York

– originally from the University of Minnesota, gone to New York – and was doing cross-

circulation with it, which was, by any analysis at all, was surely something you couldn’t sustain,

parents hooked up to children. It was just too difficult and too fraught with ethical issues and

potential tragic issues and so on. And they were busy working on bubblers at the University of

Minnesota, a very old form of oxygenation, goes back to the 19th century, for perfusion. And

there was – Melrose in London had a machine, a disk machine, and Bjorkin had a disk for

basically cerebral protection during cardiac surgery, in Sweden. He was in the Karolinska.

And Curtin came back and said the best machine is Gibbons’ machine. Everybody’s

familiar with that, a very long story. It’s kind of an anniversary this year. So they said, “Well,

get it. We’ll buy it,” and Gibbon wouldn’t sell it. Gibbon said, “Look, I’ve had disaster with it.

I had one success and three disasters. I don’t think it’s – there’s something wrong. I don’t know

what it is,” and he was very determined. And Curtin kept badgering him. And eventually, on

Christmas Eve, I think 1953 but possibly 1954; I think it was 1953, a Gibbons letter came saying,

“All right, I will agree,” and Curtin had promised him he would keep Gibbons’ name attached to

the machine. And Curtin went on Christmas day.

So he told me – now, I hope this is true, but it’s one of these things that is clear in my

memory as something that he told me – went on Christmas day to pick up the blueprints, and

they made a machine at Mayo on the basis of Gibbons’ blueprints. And they’d worked, Gibbon

had worked with IBM in Philadelphia, and IBM had come into Rochester fairly recently,


encouraged by the opportunities so it wouldn’t be a one-industry town.

And then they did ultimately, I think, ten dogs, half an hour of open heart, no procedures.

I think they opened the right atrium and closed it again. And they got survivors for long enough

that they decided it was good.

And so in March of 1955, they had started. I think the patient that I saw was about

patient number thirty-five or so, and Curtin told me how at the end of the first year he had had 25

percent mortality. That was in 1956. And the question was, would they let him go on or not?

They were fairly used to having 1 or 2 percent mortality, even in those days, from colon

resections or gastrectomies and so on. And they came out and said, “You can go on,” which was

wonderful for him, so he continued.

And the person that I was assigned to for my thesis, since everyone had to do a master’s

degree as a fellow at the Mayo Clinic. I think virtually everyone did do that. And a master’s

degree in surgery, a master of science in surgery, a very unusual degree in the United States, very

common in Britain and the British-connected places like Australia, New Zealand, India. A

master’s of surgery was an important degree, the equivalent of getting a Ph.D. in medicine. You

had to do a thesis for it.

And I said to Dr. Ellis, you know, “What should I do?” and he said, “Well, why not

develop mitral valve prosthesis.” Okay.

SJ: A master’s degree? [laughing]

RF: [laughing] And so I got to work. And I first, obviously, looked in the literature. There

was one or two people who’d been trying to do something – I forget whose name; it’s left me for


the moment – who ended up in Brooklyn – I’ll think of his name – had a metal valve with a sort

of a cage around it, and the idea was you put a purse string and pulled the tissues into this

meshwork cage, and that would then be the attachment. So that was up, and it did not work

terribly well in the few attempts at using it in dogs.

So I looked up the literature. There wasn’t much. I looked up the anatomy. There wasn’t

much. I got specimens of deer, beaver, dog, cow, sheep, a bird, various ducks, a rat, turtle, all

kinds of hearts, and I gathered them from anybody who could find me one, you know, a

gamekeeper who gave me the deer, and so on. And I looked at the way the mitral valve was

constructed and the way it worked, and I had some help from a fellow called Bowling. He was a

physiologist at the Mayo Clinic, and Grindley, who was a marvelous man. He had been an army

surgeon in Burma during the war, and he was a crusty sort of character, and he was in charge of

experimental surgery. And so I had six months, by the way, for this project.

SJ: Six months?

RF: [laughing] Six months. And so I started making designs. And what Bowling and

Grindley had done, which gave me a clue, which turned out to be an extraordinarily important

few seconds of observation, they were making hearts beat by perfusing the heart with oxygenated

Ringer’s lactate, and it was clear fluid in this beating heart. The atrium was open and, of course,

you looked through the water and all the liquid, the clear liquid. And I noticed the way the mitral

valve was beating, and I noticed that the cordi were always under some tension, even in diastole,

and I was able to get a fair idea of how the heart was working, and in particular how important

the anterior leaflet was as doing most of the movement and meeting a closed valve leaflet which


was composed of approximated scallops being forced together by the pressure of the contracting

annulus. And I took some movies of that.

And I then started to think of a variety of designs, and so I used silicone rubber for some,

and that had four leaflets, that silicone rubber valve, and I didn’t really want to make a valve that

was quarterly supported because I thought it would be too difficult.

Now, working at the same time as I was, was Nina Brownwald, Dr. Brownwald’s wife,

and she designed a polyurethane valve, mitral valve, made out of polyurethane foam in which

was incorporated a series of polyester chordal that came to two sort of thick polyester sutures and

were then pushed through the ventricle to be tightened up outside.

At any rate, we ended up with a mono leaflet valve which had flexion across multiple

hinges and stiffness parallel to those hinges in some bars of Mylar was the substance we used.

And we had used, we went out to Dow Corning, looking at various adhesives, and really that was

with Crazy Glue, actually, which we joined the stuff together with, and then made multiple

laminates to make the ring and then put on the outside of the ring a porous sewing ring.

Now, we got to the porous sewing ring by implanting partially compressed ivalon sponge

donuts above the tricuspid orifice under inflow stasis, and we showed that in three months that

porous ivalon ring – if you compress it too much, of course, it became solid, so you couldn’t. It

would have no invasion – had been invaded by fibrous tissue and covered by it as the sort of

pseudo-intima. That gave us the clue to the proper design of the sewing ring. You had to have a

porous sewing ring, one which the tissues could invade and replace the sutures.

So we made our sewing ring out of an ivalon sponge attached to the outside of the more

stiff portion of the [unclear], which was de-shaped to imitate nature, and the large and only

leaflet that opened out of the way of the blood and closed back against the ring had its straight


side on the straight side of the ring, and that was the part that was put under the aorta. So this

valve opened around the way a turtle heart opens, which is only one leaflet, or the way a mitral

valve does most of the moving.

And we did dog work, and our longest dog survived for some eighty-five days. We did

no real wear testing. We didn’t quite know how to make an accelerated wear tester. The valve

that was in there for eighty-five days looked fine when it came out. We found that there was an

enormous tendency for embolism to form. We found that clots would form on the valve

extremely quickly, build up out to the orifice, and at the orifice would, depending on the flow,

start expanding into the flow and growing into the flow, and then break off and cause an

embolism, or cross and meet and thrombose the valve. But we found that these, this particular

design could be helped by placing the residual leaflet tissue over the ring, so putting it essentially

subannular. And with that kind of insertion, we found that these dogs seemed to be surviving

without emboli.

Interestingly enough, if we put it the other way around so that it opened towards the

septum and there was in fact a dead space behind the open leaflet against the mirror annulus in

diastole, the thrombus would form behind there. Obviously, there was an area of stasis from a

slow-moving vortex that formed behind there.

And, by the way, we did test this valve in a pulse duplicator, and we had atrial ventricular

pressures and we had blue dye injected into the ventricle so we could see what insufficiency there

was and could also see, as the fluid flowed through it, what the flow pattern was like. And we

took cinnies of this.

SJ: You took what?


RF: Cinny pictures, 16mm cinematography.

SJ: I think that’s a British word. Okay.

RF: Cinnies? It’s a medical term. We use it very frequently. Cinematography is where it

comes from.

SJ: Well, that is definitely important.

RF: And Dr. Ellis said, “I think we’re ready to do patients.” So I made some adult-sized

valves up and I sewed them, sewed the ring . . .

SJ: And what year is this?

RF: This is 1960 now. My thesis started in 1959, ended in January 1960, and so now it’s June

1960.

SJ: And you’ve got the degree, right?

RF: My thesis hasn’t been awarded yet.

SJ: So you took longer than six months.


RF: Six months was the period in the lab.

SJ: The research.

RF: I’m back on the service now for the last six months.

SJ: Okay, got it.

RF: So I made up these adult valves, so, and we decided to cover them with knitted Dacron

cloth. The principle we had worked out was that we’re going to get protein deposition, starting

with fibrin for sure, certainly albumin – we knew that – within hours of putting this device in the

circulation. So long as the flow pattern was unlucky to allow for stasis, and so long as there was

a surface to which thrombus would cling, we figured that we were going to be safe from the

development of embolism.

And so we had a valve, which was a flexible, called a flexible mono cusp covered in a

very thin layer, a single layer of light knitted Dacron, and it was Dacron over the ivalon sewing

ring, and we did two patients, I think on June the 7th and June the 16th, 1960. One patient was a

Capital Airlines pilot.

Capital Airlines is long since defunct. We had traveled, in fact, on Capital Airlines from

Rochester to Dow Corning, had been in Michigan when we went to the Corning people to see

what materials they might help us with. And I knew that airline was going broke because the

airline hostesses were in carpet surplus. [laughter] Well, I said to my companion, “There’s


something funny about this airline.” Never saw airline hostesses in carpet reminants before.

Their morale must be low.

Anyway, he was a Capital Airlines pilot who had been flying with heart failure, just to

give you an idea. The FAA wasn’t as shocked as it is today. And he had actually had an attempt

at an annular prosthesis, and it had failed, so they decided he needed a valve replacement. It was

about six months after his annular prosthesis. And he left the hospital nine days after surgery

with an excellent heart sound. You really couldn’t tell the difference. His primary aortic

pressure had been very high. I still have the pressure record of the fall in left atrial pressure,

which we measured. And we did, of course, cardiac outputs with dye dilution and so on.

The second patient was a female, younger, and similarly sick person, and low cardiac

index, severe pulmonary hypertension, dominant mitral insufficiency, birth rheumatics probably

originally, and she came through fine, too, nine days after surgery. I’m sorry, it wasn’t a female.

I’m thinking of the third patient. It was a male, it was a male.

So these two patients left the hospital alive, and there had been no IRB nor, certainly, any

regulatory approval from FDA or any other body. There had been an agreement amongst the

staff, John Curtin and Bunky and, no doubt, the cardiologists that these patients had no future

without surgery, without an attempt to correct their mitral insufficiency. We know what the

prognosis is. They’ve already been in heart failure and they’ve got a limited life.

We had conscientiously tried to find out what we could about the ideas we had. Had we

been more cautious, perhaps more willing or more desirous of being absolutely certain, we, of

course, would not have started. We had to start with uncertainty.

I think, you know, I mean, one thinks about those kinds of things. I made those two

valves.


SJ: What options did they have at this point?

RF: They had the option of digitalis and mercury, mercury diuretics, digitalis. There was no

lidocaine, there was no modern type of diuretic, and digitalis was the drug of choice. There was

adrenalin to be used in the operating room, and there was even – what was the other one? –

another form of adrenalin, which I’ll think of in a second, not noradrenalin, another one. But, at

any rate, those were the agents that we had.

So, people with rheumatic fever died. They quite often didn’t die until they were in their

thirties, but generally about thirty years after making the diagnosis was about the time they died,

after it had been picked up in childhood. We didn’t really know about myopathic heart failure

then. We figured it was all due to the valve. And there were other diagnoses, which you might

recover, but, you know, you might recover from endocarditis if you had a valve lesion, but you

were going to die if that lesion was resulting in either a blocked or a leaking valve.

So they didn’t have any other options, and so we had to be prepared that we would, that

we might fail. We had to be prepared that the device might fail. And we did, in fact, tell the

patients that we don’t know what’s going to happen. It’s all we’ve got. If you want to try it,

we’re willing to try it with you.

Can we stop?

SJ: Yes.

[END OF TAPE 1, SIDE A]


RF: Three months, or just under three months after the first operation, we got a call on a

Friday night, actually, late on Friday night, that our first patient had been watching the Friday

night high school football game in Marshall Town, Iowa, and had died on the stand in front of

the hotel, and they were going to do an autopsy on him in the morning and they wanted us to

know. So we hired a plane, a Beechcraft. We flew down from Rochester to Marshall Town,

which wasn’t very far. We landed on a flying field which had corn growing on it. I remember it

very well because there was a small runway, grass runway, and the corn was higher than the

wings when I looked out as we landed.

And we went to the funeral parlor where there were a whole lot of town people – they

couldn’t all have been relatives – packed into the funeral parlor to watch the autopsy. And we

opened up the atrium, and the suture which had been used to put the valve in was 3-0 running

silk. Why had we used 3-0 running silk? Because that’s what had worked extremely well in the

normal dogs. But in this rheumatic patient with a rather scarred annular area, healing into our

sewing ring had not taken place, so – by the dog, that would be absolutely socked in in three

months of time – and the suture had broken. 3-0 silk is incredibly weak by comparison to what

we use today. And so he died of an acute disruption of the valve. The valve was intact. And so

that was the first patient.

In September, I think the 9th and 10th of 1960, there was the first meeting on artificial

heart valves. It was called Artificial Valves for Cardiac Surgery. Merindino had pulled it

together, and the NIH had financed it. It was held in Chicago. I used to think it was the Palmer

House, but I think it was not the Palmer House. I’ll have to try and look it up to see what hotel it

was in. And on Friday evening . . .


SJ: Do you still have the program?

RF: Yes. I don’t have the program. I have the book that was published from it.

On Friday evening, there was a cocktail party, as there usually is, a reception, and

Hufnagel came up to me and said, “I thought you’d be interested in these photographs. I took

them yesterday.” And they were color photographs, and I don’t know, they must have been

colored Polaroids, I think, but there wasn’t color Polaroid photography at that time. I don’t know

how he got a color photograph in less than twenty-four hours. I still have the color photograph.

And there’s his thumb in it, and he’s holding the atrium open, and it was the same picture as the

one we had seen in the autopsy.

We had gone to that conference as the only people in the world with a live patient who

had had a mitral valve replacement. And, of course, by the time we got to the conference, there

was no patient in the world alive with mitral valve prosthesis. Nina Brownwald had done the

first. Ours were the second and third to leave the hospital. We don’t know others who might

have been done elsewhere that didn’t leave hospitals. Nina’s patient, put in by Glenn Morrown,

built by her, had left the hospital. It had been several patients before then before they finally had

a success. But the success, she left, that patient left and died about three to four months later for

reasons they couldn’t determine. I don’t know whether they got an autopsy or not, but they

couldn’t determine.

Our staff had tried, about ten days before the conference, so that he, you know, he

obviously wanted to have a live patient for the conference, and that patient died of air embolism,

a massive air embolism, and that was obviously part of the learning curve of how to do mitral


valve surgery.

So there was no live patient.

SJ: But wait a minute. Your first patient died. Your second one had died at that point?

RF: That’s – Hufnagel showed us the . . .

SJ: Oh, I’m sorry. Okay. I didn’t catch that.

RF: Hufnagel showed us the picture.

SJ: And he had died the same way basically that the first one did.

RF: Yeah. And that valve continued to be put in by Ellis. It was made . . . I left in ’60,

having completed my fellowship. Sorry, I left in ’61. And the valve was made by the people at

IBM, and there were a series of them done and compared with the Starr ball valve, which they

also started using fairly soon after then, and they published this. The incidence of thrombo-

embolism was very much less. Of course, there was not anticoagulating the patients. But ours

did not have any significant emboli, whereas the Starr valve had a whole lot and they realized

they were going to have to give it, give the patients anticoagulants.

But what happened basically was that ours was an amateur effort, and even when the IBM

people were making it, it was sort of a favor and it wasn’t done for money or anything.

The great difference with Al Starr’s valve was that he had Edwards, the engineer, with


him, and everything, from the beginning, was done by an engineer, not by amateurs. And, as a

result, that valve became a success. I should add that the longest dog survivor that Al Starr had

before he put his first valve into a patient was seventeen days, and he had announced that the dog

is too difficult an animal to do valve replacement surgery, and we have to go straight to humans.

Our approach had been different at the clinic, and we got up to eighty-eight days, or eighty-five.

I can’t remember. I’ll have to look that number up. And so Al’s valve was then a success.

I went on to work on pericardium or autogenous pericardium for constructing valve

repairs, feeling that we could get a very long way with repairs rather than replace, and I had a

whole series of chordal replacements, leaflet extensions, transfer of cordi from the [unclear] to

the [unclear] when they [unclear] cordi, so we’d shorten them by taking the posterior leaf of the

[unclear] and filling in the rest with the guts of the pericardium. That, too, failed in time because

of the body’s response to autogenous pericardium.

Now, what was happening in the next five years or so was that many valve designs began

to come along. The testing was dependent on the conscientiousness of the people doing it, and

most of them were very conscientious, trying very hard, as we had been. I feel quite good about

what we were trying to do then. We were working with an enormous level of ignorance. If you

analyze what we did in the early cardiac open-heart cases, we did not have blood gases; we did

not have radial artery pressure or any other form of pressure except a blood-pressure cuff and a

finger; we did not have electrolytes; we did not have an EKG monitor. We took two paper

EKG’s a day to see what the patient’s rhythm was, and in the operating room, we looked at the

heart and analyzed it with our eyes as to whether the atrium and ventricle were in synchrony and

so on. And we would weigh our patients – they were mostly children because they had

congenital heart disease – by standing on a scale with a baby in our hands, and then knowing the


weight of the dressings and the tubes and so on, and we would control our fluid balance by

weighing the patients and watching the urine output and watching their weight. And despite that,

we actually got 100 VSD’s in succession without a mortality in 1961 with John Curtin.

I went back to South Africa and started working with Chris Barnard, who had been for

two years to the University of Minnesota, the [unclear], and in those two years got both a

master’s degree and a Ph.D. degree in surgery, which is pretty darn good, and he had come back

as the one-eyed man in the kingdom of the blind to Capetown.

We’d had a wartime thoracic surgeon, first-class pulmonary surgeon, done good work in

trauma surgery in the war, learned how to do closed mitral commissurotomies pretty well from

Brock, and thought he could do open-heart surgery by just coming on a quick trip to the United

States and buying a pump, and it hadn’t worked. Barnard made it work in Capetown even

though he was not even qualified to be a surgeon. He hadn’t done enough surgery to qualify as a

surgeon. But, as I said, in that atmosphere and with this fantastic department of cardiology, they

brought him through. I don’t want to divert too far from the subject at hand, which is the FDA,

so I don’t really. . .

SJ: Keep going. We’ve got plenty of time. Keep going.

RF: Okay.

SJ: We can finish later if we don’t finish now.

RF: The way Dr. Schirra brought Dr. Barnard on in cardiac surgery was to line up for him


twenty atrial septal defects; twenty pulmonary stenoses, the two simplest open-heart operations

that you could do; twenty VSD’s – now we’re reaching a higher level of complexity; and then,

finally, his first tetralogy of Fallot, which has a combination of ventricular septal defect and

pulmonary stenosis. And it was such a sort of extraordinarily smart thing to do for a cardiologist

not to do the conventional thing, which is give the most difficult case in a new area, a case that’s

going to die anyhow, but to give him the simplest case and have the faith in his abilities and his

ability to organize a heart-lung team, and we really were a team from the beginning in Capetown,

absolutely. We never had any sort of individualism, except Barnard’s individualism, interfering

with the working of the team.

And Barnard then got hepatitis, and he was sick. Hepatitis has been a hazard for cardiac

surgeons ever since the very beginning. Don’t forget, we used to draw fifteen units of fresh,

warm blood to do our cases at the Mayo Clinic at six o’clock in the morning, before surgery.

Anyway, Barnard announced that there would be no more cardiac surgery until I recover,

and Schirra went to see him and said, “What do you mean there’ll be no more cardiac surgery? I

want to have cardiac surgery in the University of Capetown, and we invited Frater to come back

from the Mayo Clinic,” where you got the best education he could have in those days outside of

the University of Minnesota. The two best places were undoubtedly in Minnesota, the clinic and

the university. “And he’s going to continue doing cardiac surgery because I want to have two

cardiac surgeons here. I don’t want my program to stop every time some surgeon is sick.” And

he said, “You’re too sick to argue with me.” And Barnard couldn’t argue with him.

And I heard this, the fact that the cardiac thing was going to continue, and I ran down to

Ward C3, and I saw a fellow and he had this enormous heart swelling, sort of red and angry

coming through his chest. It was an aneurysm, encephalitic ascending aortic aneurysm that was


eroding through the sternum and was under the skin, you see. And I ran to Paul. I said, “Paul,

I’m going to operate tomorrow. There’s a fantastic patient in C3.” He says, “I know that

patient.”

I said, “Well, I’m going to do him tomorrow.”

He says, “You’re not.”

I said, “What do you mean, I’m not going to do him? He’s got to be done.”

“No, he’s not going to be done.”

“He’s going to die if I don’t operate on him!”

He says, “He’s going to die if you do operate on him.”

I’ve put an ASD on the schedule for tomorrow, and you’re going to do that ASD.”

Now that, to me, was such an extraordinary act. If I think of the way cardiologists have

behaved in other areas with cardiac surgeons, he saved my life. I would have operated on this

chap, there would have been blood on the ceiling, the patient would have died on the table, and

they would have said, “Look, Barnard had a series of successes, and this guy Frater comes back

from America and thinks he’s full of it, he thinks he’s just the cat’s pajamas, and he kills his first

patient,” because Barnard had promised me a case a week, but then it had never come, and I kept

on asking, “When am I going to get my case?” “Don’t worry, you’re going to get a case, you’re

going to get a case.” It never happened. And so I was so keen to go.

Anyway, we then decided that we needed to develop heart valves too. Now, my heart

valve was essentially dead because I’d left the clinic, and the clinic had been using it, but it

wasn’t being made by anybody.

We had an extraordinary Jack-of-all-trades in our department, a fellow called Gursen or

Hursen as you’d say in Afrikaans, who could run a heart-lung machine, who could take


photographs, who could fix up anything electrical, who could work a lathe, and we started

working on another heart valve. And we in fact made a mitral valve and an aortic valve, with

silicone poppets to close them, stainless steel, machined out of a single block of stainless steel,

with a ring and the holders. Those parts that were strained or somehow contained the poppets of

the aortic valve looked like an Apollo space capsule with two sort of poles on the end of it, and it

moved up and down with two rings around those poles, so that it had round on the bottom to

meet the ring and curved off at the top. And its flow pattern, as judged by tea leaves floating in a

pulse duplicator, was absolutely smooth. The flow just came around in the smoothest

convention. And we used that valve without any coagulation, and we actually did not get emboli

with that valve. There was no area of stasis in relation to that valve at all, and there was great

washing of the junction of tissue and inorganic material with every beat.

The mitral valve was not so successful. It was a funny sort of disk on a pole with a T on

the top of the pole, and that also moved up and down in a ring extension from the sewing ring.

But those valves were made in our department by the perfusionist, barely tested in animals. This

was Barnard’s. As far as animal testing that I had done on my valve at the clinic, tested in a

pulse duplicator to show that they opened and closed appropriately, not really tested in a wear

test because, again, wear testing had not yet come along.

And the valve was a great success, worked for a long, long time, and only really stopped

when Gursen finally got tired of all the abuse he had from Chris as a routine. He always got

abuse from Chris because Chris, that’s the way he behaved in the operating room.

And once Gursen left and went to America, where he emigrated, he did extraordinarily

well and is now retired, living in Florida. Of course, it stopped being made.

So then it became necessary as time went on that you had to, you could no longer make


your own valves, you had to get a company to make them. And Barnard then started using Starr

ball valves or whatever was coming on the market. And there were many more valves coming on

the market all the time at that time.

There are now books which you can look at which chronicle hundreds – you’ve got some

of them? – chronicle the many different designs.

Clark, Dick Clark, who was I think in St. Louis at the time and ended up at the NIH,

started – and he had been an engineer – started the notion of accelerated wear testing, and another

fellow called Swanson was involved, and the two of them, not necessarily working together but

certainly collaborating sometimes, developed the notion that you could do wear testing by

speeding up the rate of opening and closing in the machine, and he began to develop some

principles of it.

And, you know, the pig valve, of course, came out of, originally from Europe, the tanned

pig valve, first Oxford with Gunning and Duran, and then Carpentier and Jean-Paul Binet

developed – Binet developed a formaldehyde-treated pig valve. Alfgaining and Carlor developed

a stent, and they were using it initially to put fascialata on it and around it to make a pretty good

valve with this rather good fibrous tissue from the thigh. I had kind of given up on pericardium

because autogenous pericardium, untreated, invariably scarred up. And so there were these

various people working in various directions.

The homograft had come in quite early, but that was so difficult to procure in many parts

of the world, and probably certainly more difficult to use. Donald Ross used it, but [unclear],

somewhat independently of Donald, started using it. Jim Malm in New York started using it.

John Curtin to some extent, but it was very small use. And, of course, no regulation on that at

all. Again, your conscience and what constitutes sterility and how can you prove sterility.


Where the aldehyde tanned pig valve came along, it had in fact virtually no control at that

time from any device agency. And Hancock and others would just sort of declare what its

properties were, but producing very little evidence to demonstrate it.

We had done a lot more work – when I say we, I mean I and colleagues in my lab – trying

to do good hemodynamic, fluid dynamic testing. And when the Ionescu-Shiley valve came

along, here is now a manufactured valve as opposed to a harvested valve, and because it was

manufactured and the pericardium was mounted around the outside of the valve, it was possible

to get better hemodynamics than the pig valves had shown. So it was largely on hemodynamic

grounds, not on durability grounds or other performance grounds, that the Ionescu valve began to

be popular.

And we then embarked on our own studies with a fluid dynamics engineer to try and

decide what constituted proper testing in fluid dynamics. I mean, there are principles which up to

this day – in fact, there’s a paper in this meeting talking about the correct way to describe the

performance of a valve is to relate the effective orifice area to the mounting size or to annular

size, they call it. We called that the performance index, and we published papers, and we

published papers on that in the ‘60s.

At the second valve conference in 1968, which was held at the Century Plaza Hotel in

Los Angeles, run by a fellow from the Mormon university, Seventh Day Adventist University –

what was his name, a tall fellow? I’ll think of it in just a moment.

And so there were individual labs, Dick Clark’s lab and Swanson’s lab, trying to do wear

testing; our lab describing the proper principles for categorizing and describing the flow through

and the regurgitation backwards through a valve, and I’m sure other people were doing it too, but

we certainly were prominent at that time. And the reason why we were able to do it was because


we had a fluid dynamics engineer who’d come to work with us, an interesting fellow. He’d lost

his job because of McCarthy, because he had been a communist. He lost his job in Seattle, and

he came, and we gave him a job, and it was a wonderful, wonderful time we had together,

[unclear] Yellin.

And then we started using, we were the first in the country to use the Ionescu-Shiley

valve because we had tested it hemodynamically against the other available devices and said it

was better hemodynamically. We were pretty unsophisticated because we didn’t really recognize

that you didn’t have to have a perfect valve to benefit the patient, but it was better. And we

published some data, some early data, on its better performance.

Now we’re talking about the ‘70s, the mid-‘70s.

SJ: Tell me a little bit – and you may be getting to this in the ‘70s; I just remember when we

had our original conversation, you had some wonderful stories about the crazy marketplace and

how some of the valves came and went without as much rhyme and reason as you might think.

RF: Well, there was always the question of, if somebody was making a valve, they had to be

selling it to be able to continue making it. The money for those ventures, I don’t really know

how it was raised, frankly. I don’t know how much venture capital was in it, I think relatively

little. I think it was more like loans and/or somebody’s independent wealth that enabled them to

start doing something. Or it was a company that had other products that were successfully being

sold, and they tried to venture into this glamorous new field.

And certainly Shiley was – Shiley was an engineer who figured he knew how to make

valves, mechanical valves. Edwards was an engineer who figured he knew how to make


mechanical valves. Both were acquired by big companies after they’d had some success as

small, starter companies. Pfizer bought Shiley and Baxter bought Edwards. Actually, American

Hospitals bought Edwards first, and then Baxter.

I’m not sure what I was talking about then. It doesn’t immediately come to me what

particular things I was saying. But let me go on a little bit.

We had our first incident with an Ionescu valve in a patient who was one year post-op

with a mitral valve in place. He was admitted to the emergency room with some chest pain, and

a relatively – we were very early into echo, and the echo showed evidence . . . No, wait a minute.

It was not echo. It was an angiogram. Sorry, I’m wrong. That case was not an echo. He had

developed a systolic murmur, so we took, we did an angiogram, a ventriculogram, and a jet of

insufficiency came out adjacent to the annulus, and we thought it was perivalvular. And two

months later, the patient came in in pulmonary edema and died, and in autopsy there was a tear

from post all the way around to the belly of the leaflet along the edge of the rigid Ionescu-Shiley

stent, along with the strut, which was cloth-covered, and we looked at it and we wondered why it

had been a jet of insufficiency at the base of the valve to start with, and we reasoned that a tear

might have occurred there and worked its way back up to the top and broken away.

Now, there was always a stitch across the leaflets at the top of each post, and it was

tempting to say, well, that stitch must have started it, but we knew this had come as a jet from the

base of the valve to start with.

So we decided we needed to do wear testing, and we set up in very quick order a wear-

testing system which we’d built ourselves. We had some guys skilled at instrumentation, as in

Capetown, we had Gursen; in New York, we had a couple of guys down in the basement who

could turn something out for us, and we built a wear tester. And we knew that we had to try and


see that it opened and closed with each beat, and we used a strobe for that. And we didn’t really

know how we could measure the pressures at such high frequency, and we did eventually learn

how to do it. But we started on testing of a series of Ionescu valves, and we found, as we

watched them in the machines, going at 1,800 cycles a minute, which is far faster than anybody

tests a valve now, that tears would develop commonly along the belly of the leaflet close to

where, right at its contact with this fabric that tore, occasionally up at the top, but it would then,

from a small tear, would spread. And we reasoned by observation that when you saw there was

contact with the cloth, and we reasoned that this was an abrasion wear problem. And we

published a paper on that.

And, as a result of it, of course, we stopped using the valve. And here’s where the need

for control came in. Because we told Marian what was happening, the Romanian surgeon in

Leeds who invented the valve, Marian wasn’t 100 percent straightforward with us. He said he’d

never seen it, and maybe he hadn’t. And we approached Shiley about it, and they said we didn’t

know how to put a valve in. And we stopped using it and were pretty damned sure we had

established the cause. And as we went on following our patients, we found other patients

developing the complication – almost all mitrals, almost all big mitrals, where the relationship

between sickness of pericardium and size of the valve and stress on the closed leaflets was highly

likely to be acting rather more easily in the big valves than in the small valves. We also figured,

we learned something about procurement . . .

[END OF TAPE 1, SIDE B]

RF: We learned that pericardium was being harvested from a variety of slaughterhouses


around the country, put into saline, and trucked across the country, and that it could be forty-eight

hours at relatively high ambient temperatures that the pericardium had been in transport. We

recognized that there was great variability in the time at which tears would occur, and we

reasoned that it was possibly because of irregular procurement or insufficient quality control of

the pericardium while in transport and at the factory, and perhaps, perhaps variations in

autodigestion and bacterial contamination of that pericardium before it got sterilized.

And then, finally, we also thought that since this was a handmade device, and you can

make, a bad tailor can make a pair of trousers that tears as soon as you bend over or shift your

shoulder or something, that some of the sewers may have had more contact in the valves they

produced with the leaflet material and the club of the ring than others, because there was clearly a

difference. And so we published some material on this.

And we then were offered, because we had become people who’d used the Ionescu-

Shiley, offered the use of a series of other new valves made out of pericardium. Hancock made a

valve, Ionescu came out with the Ionescu 2, and in meetings with Marian, he told me how he’d

been trying very hard to make the Shiley Company put a pericardial covering rather than a cloth

covering on the frame of the stent. And, indeed, they had in-house data which showed that

pericardium-covered stents did not have the breakage rate.

When I asked them what they had done in terms of testing beforehand, they of course said

we hadn’t tested it, because there was no way to test it beforehand, before it came on the market.

Our testing was one of the very first papers to make a systematic effort to define something like

this.

SJ: And what year was that published, roughly? We can check.


RF: Seventy something, ’72, something like that. I could get you my bibliography, my CV

from my secretary. It’s got those things all buried in there somewhere.

And Marian told me of his great frustration because Bruce Fitell, the CEO of Shiley, was

not interested in making a change until they’d gone through the whole process of now they were

beginning to develop a testing program. He didn’t want to do it, and there was reluctance on

Bruce’s part, and I think the reluctance was related to admitting there was something wrong with

the valve, which he never did admit, and taking care of inventory, you know, that sort of thing.

So, that may be an unfair characterization, but it’s the impression I got, and Marian was

frustrated, there was no argument about it.

And then we tested in our wear tester. At this time, the late ‘70s by now, when new

valves are coming into use and under test; maybe it was by the mid-‘70s before we . . . No, it

must have been the late ‘70s, sorry, that we did our testing, not ’72, late ‘70s. And now it was

early ‘80s, about 1980 or so, where we were testing a new valve produced by Edwards. The

Hancock 2, the Ionescu 2, and the Mitroflow.

The Hancock 2 was horrible in terms of its wear testing, fantastic with hemodynamics,

the best hemodynamics and the worst wear testing.

The Ionescu 2 was only marginally better hemodynamically, if at all, than the earlier

Ionescu, and very poor in wear testing.

The Edwards was reasonably good on wear testing and fair on hemodynamics, not as

good as the Hancock 2, but it was certainly better than the Ionescu and certainly better than the

concurrent pig valve would have been, or was, as we tested them.

And the Mitroflow was not too far off the Edwards in durability.


And, mind you, all of these were not lasting 100 million cycles. They were failing much

before then. And I think our system was excessively destructive at the rate we were running, and

we really began to learn how to do pressures. We were able to get instantaneous, brief, the

spikes of pressure, and we knew what the spike, very short-lived spike of systolic pressure was in

these valves.

SJ: From patients.

RF: No, from the machines, from our machines, you see. We were developing . . .

SJ: The machines could measure that.

RF: There was no guidance to tell us what we should do.

And, anyway, we chose to use the Edwards, and not in the mitral position. We chose to

use it only in the aortic position and told Edwards so. And so we were then in the trial, the FDA

trial of the Edwards valve, from ’82 to ’84.

And, at the same time – I’m trying to remember.

Anyway, from our point of view, we had developed a way of looking at valves, and AMI

at that time started – and I was on the valve standards committee, the valve committee – and we

started to develop a sort of a template for testing valves in vitro and began to approach the animal

testing, where we really never knew what good animal testing would be. The dog was clearly not

a very good model. We hadn’t come upon the sheep yet because the one who did that was Jones

at the NIH.


SJ: Why was the dog not a good model?

RF: It gets bacterial endocarditis extremely easily. It gets bacteremia far more commonly than

other animals do. I guess carnivores may get bacteremia more easily, I don’t know, than

ruminants, but it was extremely prone to endocarditis. Quite a good model to operate on, quite

nice anatomy, but not at all good to look after long term.

I’m trying to think of when the Pfizer, the Bjork-Shiley thing, began to blow up,

sometime in the late ‘70s, I believe, and early ‘80s.

SJ: Perhaps ’75 to ’83-ish.

RF: The point being that Bruce Fitell behaved in exactly the same way with the Shiley as –

with the Bjork-Shiley as he behaved with the Ionescu-Shiley. A couple of . . .

You know, the valve had been working well. It had thrombosed from time to time with

poor anticoagulation and would thrombose in the lesser orifice, and the Shiley people recognized

this and therefore changed the design with a circle CC, concave and convex, to make it – they

shifted the axis close to the center, and so the secondary orifice became bigger. And it became

very quickly apparent that the incidence of thrombosis had clearly dropped – not the incidence of

embolism, which is a different phenomenon from thrombosis. Embolism comes from the

interface at the edge of the valve between tissue that’s grown there and inorganic material. If

there’s any kind of stasis area, then these little different aggregates build up and break away. But

thrombosis of the valve is related to a sort of red clot forming in an area of stasis relative to the


valve, and that was much better with the new Bjork-Shiley.

And then the strut, the output strut fractures began to occur, and it was very evident that

Bruce knew about them, Bjork surely knew about them. Bjork was trying very hard – and I’m

sure you’re familiar with the famous letters that were written – trying very hard to get them to

change. And I spoke to Bruce, and the FDA – this was my first knowledge of an FDA group

beginning to interact, not with a guidance, not with an attempt to set standards, but with

complaints that came to their ears and which made them say, “There must be something. Can

you tell us about it?” And Bruce Fitell’s response was, “They’re a bunch of jerks that don’t

know what they’re talking about, and I don’t have to listen to them at all.”

And it wasn’t the FDA that closed Bruce down. I think I can probably fairly say that. It

was the fact that it became so much public knowledge that this valve. was failing, and there was

such an outcry from the people following patients and trying to analyze them. And tell me if I’m

wrong, please, that the company realized they had to stop, and Bruce realized that he had to, he

could not possibly resist the FDA’s efforts to get information anymore, and the company realized

that they had such a hot potato in their hands that they just quit. They kept on with the new

design, which was already well advanced by then and which Bruce did not want to introduce till

it was ready to go because he wanted to have uninterrupted sales. And I think I can say that with

fair accuracy. I don’t believe that’s an incorrect statement. And anyway, he brought Shiley

down. The company ceased to function after that.

Pfizer recognized they had a bunch of losses on their hands, and the litigation lawyers

were just beginning to understand the concept of a class-action suit and what they could do with

a class-action suit. And Pfizer stopped because it was a terrible investment and it was going to

lose them money if they didn’t stop.


But I believe it was Bruce Fitell’s intransigence, his arrogance, and his behavior towards

the FDA that made the FDA recognize that they had to start making a law about this. And ’86

was the year in which they first . . .

SJ: Seventy-six.

RF: Sorry, ’76, right. Not ’86, of course not. Seventy-six was the time when they brought

that in.

So Shiley had had sort of two valves fail in fairly short order, one of which, neither of

which could necessarily have been anticipated, although if the Bjork-Shiley people had done

proper wear testing, which they did not do when the first strut failures occurred, and

subsequently, of course, they did an immensely, a rather good piece of work that essentially

worked for Shiley, who’d been a fluid dynamics engineer himself, and worked for Shiley in the

sort of, during the bankruptcy period and during the litigation period, getting to the bottom of the

cause.

SJ: Did you ever do any wear testing of the Bjork-Shiley?

RF: No, no. I never did any.

SJ: Didn’t use it.

RF: I just knew. I didn’t use – I used the Shiley once or twice maybe. But I never was a


regular user of it.

After I stopped using the Capetown valves, I did briefly use a valve – what was his name?

– a valve which is a Teflon, a caged disk valve, Teflon-covered cage and a Teflon disk. What

was his name from Texas?

FJ: Beall, Bell?

RF: Beall, Beall, the Beall valve, which indeed wore out, and that was a relatively brief

period. I never used a Starr valve. I think I put one Starr valve in in my life, probably because I

frankly thought it was hemodynamically not all that good and, in the mitral position, tended to

poke across and hit the septum if it was anything but a much dilated ventricle and give

ventricular arrhythmias.

And then I used the, started to use the St. Jude valve until St. Jude was sued by Carver

Medics for trying to make their own pirated carbon, at which point I went to use the Medtronic

Hall valve, and I used the Medtronic Hall valve for a good number of years, until it became

apparent that there were hazards in the use of it in terms of suture entrapment, and at that point

St. Jude was back on track, so I went to the St. Jude valve after that for a mechanical valve.

FJ: Was the trouble with the Hall Medtronic in terms of insertion?

RF: If you have – a suture can catch on the hooks in a mitral insertion very easily. A suture

can catch, and they can catch but you can’t see them. The hook is deep. A suture can catch on

the invisible hook of the, in the aortic position during the closure of the aortotomy. I saw that


happen once. I saw the opposite happen. And I saw a most weird one once where a suture had

been caught around the hook that was invisible in the mitral position in line with the axis of the

valve, so when you tilted the valve to look behind it, you couldn’t see the suture, but which, at

one point, had contact with the disk every time the valve closed. And a month after surgery, it

broke, and that long piece of suture came away, of course, from the hook and got into the orifice,

and as soon as it got into the orifice, the valve jammed tight shut.

And it was one of my patients, and I’d gone down – it was a coroner’s autopsy. Some kid

died at home. It was a young woman; she’d died at home. And I went down to the coroner and

insisted that he examine the heart, because otherwise he’d never know why this patient had died,

and we found this valve jammed by a suture.

And you know what the sad thing about that was? I was going to publish it, but the

lawyers, the defense lawyers for the hospital, refused to let me publish it. And, of course, they . .

. The case was ultimately thrown out, I think, and it’s still sitting there waiting to be published.

It’s one of those half-written manuscripts that never sees the light of day. You must have seen

them. [laughs]

FJ: Oh, yes. You’ve given us quite a bit, and it is getting close to six o’clock, so I want to

honor your time as well.

RF: Fine, thank you.

Male: And you wanted to go over to your hotel before you went back.


RF: Wonderful. I think I really would like to. Look, I’ll be happy to meet you anytime, you

know. Clearly . . .

SJ: This is a good thing to get started on in understanding the history of heart valves and their

regulation.

RF: My most important relationship with standards was related to AMI, you see, and the fact

that we were trying to develop standards before the FDA, I think, had really developed standards

of their own. And then ultimately, of course, there was cooperation years later between [unclear]

and other European agencies and FDA to try to get to some harmonized standard, and I think

with a bit of luck, that’ll happen.

But it’s still, unfortunately, a very difficult area because the testing methods are still

uncertain. And you know, for example, the Copper Medics oxidation-treated pericardial valve,

which tested beautifully in the tester, wear tester, but subsequently realized was probably not

moving through a full cycle because it was so much more, or through a cycle that allowed the

potential for abrasion wear to occur, because it was so pliable, it sort of floated out of the way,

and then that valve, fifteen years after the Ionescu valve had had abrasion-wear test, developed

abrasion wear. You’d think by now, my God, we should never design another pericardial valve

that could have abrasion wear, but it happened.

It’s a very interesting subject. We’ve gone from primitive absence of any form of proper

testing to serious efforts at doing a job of testing, but largely individual efforts without agreement

on what should be a standard.

And with the fluid dynamics, it’s very interesting. The FDA at one point had a series of


standard valves that they passed around to ten or eleven different labs to do fluid-dynamic testing

in vitro, and that you should be able to get to. It’s sort of an art, but it is reproducible. Every

single lab had a different answer.

SJ: That’s interesting and I will study . . .

RF: And that was an effort by FDA to get some science into this thing.

SJ: Well, thank you so much.

[END OF INTERVIEW]

oral history interview with robert frater cardiothoracic

Documents